Project Guidance

Max Pellert (https://mpellert.at)

Deep Learning for the Social Sciences

General

The course assessment is based on the assignment submissions and a final course project.

Projects are done in groups that can have - at most 4 students - must mix students with different backgrounds (at least two different backgrounds per group).

The purpose of the project is to show what you have learned in the course in terms of the basics of neural networks, data management, analysis and interpretation.

General

The lecturers will propose projects that start from a well-defined social scientific research question with an empirical focus using a technique from deep learning covered in the course.

At the same time, participants should critically reflect on those methods, what conclusions they can take from them and their limitations.

Deadlines

The timeline of the project deadlines can be found below. The deadlines are up to the end of the day in Germany marked by each date.

  • 04.06.2024: Assignment of project topics.

  • 02.07.2024: Project presentations session

  • 27.08.2024: Deadline for submission of final report.

Project Steps

1. Form a group

Get in touch as early as possible with fellow students over the course discord server or around lectures to form a group. You can use the same group composition as in former courses, but you don’t have to.

2. Project assignment

The lecturers will prepare detailed project task descriptions for each of the project groups and assign one of them to each group. You can check out an example project already now on GitHub.

Project Steps

3. Project presentations

During project presentation sessions, students give 5-minute presentations about their projects. This is an opportunity to get feedback, by presenting four slides:

  1. Short recap of the assigned research question and methods
  2. Work plan, assessment of potential problems or problems already encountered
  3. Preliminary results
  4. Conclusion and critique

Project Steps

4. Submit the final report

Send a final report as a PDF document (max. 6 pages, min. font size 11pt) via email (max.pellert@uni-konstanz.de & giordano.de-marzo@uni-konstanz.de).

Project reports should follow this structure:

  1. Data Presentation and Preliminary Analysis
  2. Deep Learning Model Implementation
  3. Research Analysis

Project grading

  • 10% for the presentation
  • 90% for the final report (30% for each of the 3 thematic blocks of the report)

Projects do not need to report “positive results”, what is important is that you show how you have shown visibly credible effort to address your research questions, document any issues or deviations, and critically reflect on methods and results.

Remember for the overall grade, that on top of the project presentation and report (60%) you can individually achieve an additional 4x10% (up to 40%) with handing in the assignment solutions.

Example project task assignment

Take a look at “ProjectExample_DLSS.md” on GitHub

Please make sure to go through the more detailed description in “ProjectsGuide_DLSS.md”

Questions?